VGG块
定义了一个名为vgg_block的函数来实现一个VGG块
import torch
from torch import nn
from d2l import torch as d2l
def vgg_block(num_convs, in_channels, out_channels):
layers = []
for _ in range(num_convs):
layers.append(nn.Conv2d(in_channels, out_channels,
kernel_size=3, padding=1))
layers.append(nn.ReLU())
in_channels = out_channels
layers.append(nn.MaxPool2d(kernel_size=2,stride=2))
return nn.Sequential(*layers)
VGG网络
与AlexNet、LeNet一样,VGG网络可以分为两部分:第一部分主要由卷积层和汇聚层组成,第二部分由全连接层组成。
VGG神经网络连接的几个VGG块(在vgg_block函数中定义)。其中有超参数变量conv_arch。该变量指定了每个VGG块里卷积层个数和输出通道数。全连接模块则与AlexNet中的相同。
原始VGG网络有5个卷积块,其中前两个块各有一个卷积层,后三个块各包含两个卷积层。 第一个模块有64个输出通道,每个后续模块将输出通道数量翻倍,直到该数字达到512。由于该网络使用8个卷积层和3个全连接层,因此它通常被称为VGG-11。
conv_arch = ((1, 64), (1, 128), (2, 256), (2, 512), (2, 512))
实现VGG11
def vgg(conv_arch):
conv_blks = []
in_channels = 1
# 卷积层部分
for (num_convs, out_channels) in conv_arch:
conv_blks.append(vgg_block(num_convs, in_channels, out_channels))
in_channels = out_channels
return nn.Sequential(
*conv_blks, nn.Flatten(),
# 全连接层部分
nn.Linear(out_channels * 7 * 7, 4096), nn.ReLU(), nn.Dropout(0.5),
nn.Linear(4096, 4096), nn.ReLU(), nn.Dropout(0.5),
nn.Linear(4096, 10))
net = vgg(conv_arch)
构建一个高度和宽度为224的单通道数据样本,以观察每个层输出的形状。
X = torch.randn(size=(1, 1, 224, 224))
for blk in net:
X = blk(X)
print(blk.__class__.__name__,'output shape:\t',X.shape)
Sequential output shape: torch.Size([1, 64, 112, 112])
Sequential output shape: torch.Size([1, 128, 56, 56])
Sequential output shape: torch.Size([1, 256, 28, 28])
Sequential output shape: torch.Size([1, 512, 14, 14])
Sequential output shape: torch.Size([1, 512, 7, 7])
Flatten output shape: torch.Size([1, 25088])
Linear output shape: torch.Size([1, 4096])
ReLU output shape: torch.Size([1, 4096])
Dropout output shape: torch.Size([1, 4096])
Linear output shape: torch.Size([1, 4096])
ReLU output shape: torch.Size([1, 4096])
Dropout output shape: torch.Size([1, 4096])
Linear output shape: torch.Size([1, 10])
训练模型
# 由于Fashion-MNIST数据集比较简单,缩放前面几个块的通道数,足够训练这个数据集
ratio = 4 # 设置比例
small_conv_arch = [(pair[0], pair[1] // ratio) for pair in conv_arch]
net = vgg(small_conv_arch)
X = torch.randn(size=(1, 1, 224, 224))
for blk in net:
X = blk(X)
print(blk.__class__.__name__,'output shape:\t',X.shape)
Sequential output shape: torch.Size([1, 16, 112, 112])
Sequential output shape: torch.Size([1, 32, 56, 56])
Sequential output shape: torch.Size([1, 64, 28, 28])
Sequential output shape: torch.Size([1, 128, 14, 14])
Sequential output shape: torch.Size([1, 128, 7, 7])
Flatten output shape: torch.Size([1, 6272])
Linear output shape: torch.Size([1, 4096])
ReLU output shape: torch.Size([1, 4096])
Dropout output shape: torch.Size([1, 4096])
Linear output shape: torch.Size([1, 4096])
ReLU output shape: torch.Size([1, 4096])
Dropout output shape: torch.Size([1, 4096])
Linear output shape: torch.Size([1, 10])
lr, num_epochs, batch_size = 0.05, 10, 128
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size, resize=224)
d2l.train_ch6(net, train_iter, test_iter, num_epochs, lr, d2l.try_gpu())
loss 0.181, train acc 0.933, test acc 0.919
378.9 examples/sec on cuda:0
· 本文使用了大量d2l包,这极大地减少了代码编辑量,需要安装d2l包才能运行本文代码
封面图片来源
欢迎点击我的主页查看更多文章。
本人学习地址https://zh-v2.d2l.ai/
恳请大佬批评指正。